Residential choice profiles, rail traffic exposures and cardiometabolic health in Southwest Sweden: explorative latent class analysis for studying inequities in environmental health

Environmental exposures might contribute to health inequitiesamong groups with varying social vulnerability. We aimed to conceptualize and characterize residential choice as an indicatorof social vulnerability and to examine its relationships with railway exposures and cardiometabolic outcomes. This exploratory theory-informed (i.e. intersectionality; social practice) latent class approach to residential choice highlights unequal exposures and outcomes. In this cross-sectional study (2017), we surveyed 7,280adults living within 1 km of trafficked railways in Sweden. Modelled railway noise and vibration were linked to residential addresses and combined with registry-based diabetes and hypertension data. Weight, height and waist circumference were self-reported. Using latent class analysis, we identified five profiles based on socioeconomic, household, structural, and lifestyle dimensions. Differences in exposure levels and health outcomes across profiles were assessed using hypothesis tests and regression models. Exposures and healthoutcomes differed significantly by profile with higher estimates in Class C (middle-aged, rural) vs. Class D (middle-aged, high-income). This applied to noise (β = 2.3 dB, 95%CI 1.78–2.82) and vibration (prevalence of ≥0.3 mm/s 10% vs. 4.7%, p < 0.001). Obesity and central obesity were more prevalent in Class C vs. D (20.9% vs. 13.2%, OR 1.78, 95%CI 1.41–2.25; 41.1% vs. 35.4%, OR 1.36, 95%CI 1.14–1.62). Diabetes and hypertension were most common in Class E (older/retired) vs. Class D (14.4% vs. 3.7% and 49.9% vs. 16.2%, p < 0.001). Still, Class C had higher odds of hypertension than class D (OR 1.44, 95%CI 1.11–1.87). Findings illustrate the value of combining theory- and data-driven methods to advance environmental health equity research.